Show HN Earned vs. Burned, Claude skill for measuring AI delivery value (github.com)

A new open-source Claude skill has been introduced for measuring impact in AI-native software delivery through an innovative "Earned vs. Burned" framework. This tool is designed to address a common issue in AI development: the reliance on traditional metrics like token usage and lines of code, which fail to capture genuine business value. Instead, it emphasizes the importance of tangible, verifiable outcomes, encouraging teams to shift their focus from mere efforts to actual earnings that correspond to business impact. Significantly, this framework allows teams to evaluate their tasks across a five-level hierarchy that gauges progress from ‘Not Started’ to ‘Outcome Verified.’ The skill can integrate with popular project management tools, calculate key metrics such as Earn Rate and Earned per AI Token, and generate concise reports that better demonstrate the value of AI projects. With the potential to enhance delivery culture by encouraging teams to always consider the final outcomes, this Claude skill offers a transformative approach to measuring success in the AI/ML community and sets a new standard for productivity evaluation.